Fracture identification and evaluation method

Abstract

A fracture identification and evaluation method belongs to the field of oil-gas exploration, and comprises the following steps: 1, setting window length and step length parameters, and reading a No.1 polar plate azimuth angle of a multi-arm dip angle logging instrument and conductivity data of all polar plates; 2, regularizing azimuth angles of other polar plates in a range of 0-360 degrees by taking the azimuth angle of the No.1 polar plate as a reference; 3, constructing a well periphery azimuth conductivity two-dimensional matrix M (xi, yi) within the window length, and taking the azimuth angle and the conductivity as xi and yi coordinates of the matrix respectively; 4, performing two-dimensional matrix color code mapping to generate a simulated image; 5, checking and optimizing the image, and judging whether the image effect meets the expectation or not; and 6, identifying the fracture type through image visualization, and quantitatively calculating six parameters. The problems that a conventional well logging interpretation method is poor in visualization effect and difficult to quantify, and an imaging well logging interpretation method is low in data collection rate and has a blind area in an image are solved.

Description

Technical Field

[0001] The invention belongs to the field of oil and gas exploration, and specifically relates to a fracture identification and evaluation method. Background Art

[0002] Faults refer to cracks or discontinuous interfaces in rocks, including two types: faults and cracks. China has a large number of small landmasses and has long been affected by the three major plates of Siberia, the Pacific Ocean and India. It has undergone multiple periods of tectonic movements, strong collisions and compression, resulting in relatively developed faults in most sedimentary basins in China. With the improvement of oil and gas exploration and development, people are paying more and more attention to the control of faults on oil and gas migration and accumulation. However, faults have diverse morphologies, complex causes and distribution patterns, and fault identification and evaluation are industry-wide problems. Existing methods have low accuracy and high costs, making it difficult to support systematic research on the impact of faults on oil and gas migration, accumulation and storage, resulting in insufficient utilization of oil and gas layers and waste of oil and gas resources.

[0003] At present, the commonly used fault identification and evaluation methods mainly include geological outcrop and core sample observation method, dynamic test method, geophysical data interpretation method, etc., which are restricted by factors such as data acquisition rate, oil and gas well type and timeliness. The three methods have different application characteristics. The geological outcrop and core sample observation method is a more traditional fault research method. Through systematic sampling observation and quantitative testing of field geological outcrop profiles, drilling cores, rock thin sections, etc., the length, height, density, dip, strike, and inclination of the fault are analyzed. It has typical advantages such as intuitive observation, quantitative testing, and high reliability. However, geological outcrops are mostly distributed in the structural erosion area at the edge of the basin, and drilling coring is only carried out in a small number of key target layers of exploration wells. Rock thin sections are mostly prepared by local sampling of cores, resulting in low data acquisition rate and weak contrast, and large-scale application is subject to certain limitations. The dynamic test method is to inject water, gas or other tracers into oil and gas wells, take samples and monitor in surrounding wells, analyze the samples, obtain the output of tracers, clarify the advancement direction, displacement speed, affected area, etc. of the tracers, so as to reflect the connectivity of the strata between oil and water wells during water injection development. This method is mainly used in the evaluation of oil well water injection effect, residual oil distribution, development policy adjustment, fracturing effect monitoring and other studies. It can indirectly reflect the scale, direction and length of the fracture, but cannot directly identify the fracture, and it is difficult to quantitatively evaluate the fracture length, height, density and other parameters. It also has the limitations of high cost and poor timeliness. The geophysical data interpretation method is to process and interpret geophysical logging, seismic and other data, use geological outcrops or core samples for calibration, summarize the fracture response characteristics of geophysical data, and then quantitatively identify and evaluate fractures. Using seismic data, methods such as ant tracking fault prediction, seismic attribute fault identification, multi-scale sweet spot fault prediction, and BP neural network fault identification have been developed. However, the vertical resolution of seismic data is mainly 8 to 15 meters, and there are limitations such as sporadic collection and local measurement.

[0004] Geophysical logging data has unique advantages such as universal detection of oil and gas wells, strong contrast and high precision. In-depth research on logging identification and fault evaluation methods has important theoretical and practical significance.

[0005] The existing conventional logging data fault identification methods mainly include three methods: conventional logging curve manual experience identification method, conventional logging curve quantitative identification method and formation dip logging interpretation method. The traditional method of conventional logging fault identification is to manually replay the logging curve and analyze the change characteristics of each conventional curve to infer the fault development zone. For example, when different types of faults are developed in the formation ( Figure 1 ), drilling mud invades the formation along the fault, causing the resistivity logging curve to change frequently ( Figure 2 ), the acoustic time difference curve jumps sharply ( Figure 3) and other response characteristics. However, this method has poor visualization effect, is difficult to quantify, has a large number of factors affected by human subjective judgment, has strong multi-solution, and poor repeatability. Conventional logging data quantitative identification methods include probability index fault identification method, information fusion fault identification method, wavelet transform fault identification method, neural network fault prediction method, etc. However, these methods are often random when selecting modeling samples. At the same time, the modeling samples directly determine the calculation results. However, it is difficult to select generally representative data samples, or they rarely exist. As a result, the above methods have good application effects in the laboratory, but the effects are uneven in the field promotion practice of oil and gas fields. The visualization effect is still poor and it is difficult to quantitatively evaluate the fault parameters. The formation dip logging interpretation method uses a multi-arm formation dip logging instrument (commonly used instruments include four arms, six arms, eight arms, sixteen arms, twenty-four arms, etc., taking six arms as an example) to measure the azimuth of the No. 1 plate (the angle between the No. 1 plate and the due north direction) and the conductivity (or resistivity) data of the No. 1 to No. 6 plates under different depth conditions by rotating and lifting from bottom to top in the wellbore. The instrument can accurately measure formation data information under complex conditions such as irregular wellbore and non-centering of the instrument through the six-arm electrode plate. At present, based on the formation dip logging data, fracture identification methods such as conductivity anomaly detection, wellbore comparison, and conductivity comparison have been established. This method can be used to identify and evaluate a single medium- and high-angle fracture with a large opening, but it is difficult to identify and evaluate low-angle fractures and induced fractures, and the visualization effect is still poor. Therefore, conventional logging methods have poor visualization effects and it is difficult to quantitatively evaluate parameters such as fracture length, height, density, dip, strike, and tendency.

[0006] The imaging logging interpretation method is currently a relatively reliable method for fault identification and evaluation, but it still has two shortcomings. First, the data acquisition rate is extremely low, and it cannot be implemented in cased wells, which limits its promotion and application. Second, affected by the gap between the measuring plates, the imaging coverage rate is mainly 60% to 80%, and there are white strip blind areas in the imaging image, which makes it difficult to support the system to carry out fault identification and evaluation research. Summary of the invention

[0007] The purpose of the present invention is to propose a method for fault identification and evaluation, which solves the problems of poor visualization and difficulty in quantification of conventional logging interpretation methods, low data acquisition rate of imaging logging interpretation methods, and blind spots in images.

[0008] The technical solution adopted by the present invention is:

[0009] A fracture identification and evaluation method comprises the following steps:

[0010] Step 1: Set the window length parameters and step length parameters, and read the azimuth of the No. 1 plate and the conductivity data of all plates of the multi-arm formation dip logging tool;

[0011] Step 2: The azimuths of each adjacent plate of the multi-arm formation dip logging instrument are separated by 360° / n, where n represents the number of plates; taking the azimuth of plate No. 1 as a reference, the azimuths of the other plates are regularized within the range of 0° to 360°, that is, when the azimuth of any plate is greater than 360°, the azimuth of the plate is subtracted by 360°, so that the azimuths of each plate are between 0° and 360°;

[0012] Step 3: Construct a two-dimensional matrix M(x i ,y i ), first divide the 360° azimuth of a certain depth point within the window length into m×n equal parts, where m represents the number of equal parts between two adjacent plates, and m≥n; use the azimuth of each plate to calculate the azimuth of each equally divided point as the matrix M(x i ,y i ) i Coordinates, and at the same time, the conductivity of each equally divided point is calculated by using the azimuth angle and conductivity data of each plate as interpolation sample points, as the matrix M(x i ,y i ) of y i coordinate;

[0013] Determine whether the data within the window length depth range has been processed. If not, increase the depth by one step, jump to the next sampling depth, and execute steps 2 to 3 until the azimuthal conductivity matrix within the window length depth range is fully constructed;

[0014] Step 4: Perform two-dimensional matrix color mapping to generate a simulated image. First, set any 2 to 5 color spectra with different RGB color modes, and map the two-dimensional matrix M(x i ,y i ) According to the conductivity value y i Arrange in order from small to large, when the conductivity values ​​are equal, according to the orientation value x i Arrange in order from small to large. When the conductivity value and azimuth value are equal, they are sorted in the same order. Then perform chromatographic partitioning to generate a simulated image of the azimuth conductivity around the well. The image coverage rate can reach 100%, which is significantly higher than the coverage rate of traditional electrical imaging logging images.

[0015] Step 5: Image inspection and optimization, determine whether the image effect meets expectations. If not, check whether the parameter settings are reasonable, reset the parameters, and execute steps 1 to 5 to optimize the image until the effect meets expectations. The parameters that need to be checked include window length, step length, number of azimuths, and color spectrum. After that, determine whether the data of the entire well section has been processed. If not, increase the step length at a deeper level, jump to the next window length, and execute steps 1 to 5 until the data of the entire well section has been processed.

[0016] Step 6: Identify the fault type through image visualization and quantitatively calculate six parameters, namely fault density, length, height, dip, strike, and inclination, and then save the data and images.

[0017] Furthermore, in step one, the multi-arm formation inclination logging instrument is selected from one of four-arm, six-arm, eight-arm, sixteen-arm and twenty-four-arm formation inclination logging instruments.

[0018] Furthermore, in step 1, the window length parameter takes a value of 1.0 to 10.0 m; and the step length parameter takes a value of 0.00254 to 0.001 m.

[0019] Furthermore, in step three, the interpolation method used in the interpolation calculation is selected from one of four interpolation methods: linear interpolation, cubic spline interpolation, cubic B-spline interpolation and Akima interpolation.

[0020] Furthermore, in step 4, the color spectrum is set to four colors: white, yellow, red, and black, and is divided into three partitions: white-yellow, yellow-red, and red-black.

[0021] Further, in step 4, the chromatogram partitioning process is: the matrix element at the first position in the sorting is corresponding to white RGB (255:255:255), the matrix element at the third position in the sorting (if the position is not divisible, the position is determined by rounding) is corresponding to yellow RGB (255:255:0), the matrix element at the second third position in the sorting (if the position is not divisible, the position is determined by rounding) is corresponding to red RGB (255:0:0), and the matrix at the last position in the sorting is corresponding to black RGB (0:0:0), thereby dividing the two-dimensional matrix into three chromatogram partitions;

[0022] The RGB color scale values ​​corresponding to the matrix elements of the remaining sorting positions are obtained by the mixed operation mapping method to realize the well circumference azimuthal conductivity two-dimensional matrix M(x i ,y i ) is mapped one by one to the RGB mode color scale, as follows:

[0023] ① Determine the multiplication factor K and addition factor A of the mixed operation mapping method: Assume that the minimum conductivity value within each chromatographic partition is E min , the maximum conductivity is E max , the left color scale value of each chromatographic partition is C lvalue , the right color value is C rvalue , when E min =E max When the multiplication factor K = 0, the addition factor A = C lvalue ; when E min ≠E maxWhen , the multiplication factor K and the addition factor A are calculated using formula (1) and formula (2) respectively:

[0024]

[0025] A=C lvalue -K×E min (2)

[0026] ②Use the mixed operation mapping method to achieve a one-to-one correspondence between the two-dimensional matrix and the color label: Let the two-dimensional matrix M(x i ,y i ) color scale corresponding value is M RGB (x i ,y i ), then we have formula (3):

[0027] M RGB (x i ,y i )=M(x i ,y i )×K+A (3)

[0028] ③ After completing the two-dimensional matrix color code mapping of the first chromatographic partition, go to the second chromatographic partition and execute steps ① and ② until the two-dimensional matrix color code mapping within the range of all chromatographic partitions is completed.

[0029] Beneficial effects of the present invention:

[0030] The present invention utilizes the azimuth of the No. 1 electrode plate of a multi-arm formation dip logging instrument and the conductivity (or resistivity) data of all the electrodes, and implements five processing steps, including setting parameters and reading data, regularizing the electrode plate azimuth, two-dimensional matrix of structural azimuth conductivity, chromatogram setting and two-dimensional matrix color code mapping, image inspection and optimization, as well as fracture identification and parameter evaluation, to realize visual identification of fractures using conventional logging data and quantitative evaluation of six parameters, including fracture density, length, height, dip, strike, and tendency.

[0031] This method combines the advantages of high acquisition rate of conventional logging data and strong visualization of imaging logging, mines the implicit information of conventional dip logging conductivity and plate azimuth data, establishes azimuthal conductivity imaging method for identifying and evaluating fractures, performs digital image processing on conventional logging data, and can intuitively identify high-angle, medium-angle, low-angle fractures and induced fractures. Compared with conventional logging interpretation methods, it has the advantages of strong visualization, high reliability and quantification; compared with imaging logging interpretation methods, it has the advantages of universal survey of oil and gas field data, low cost and high imaging coverage. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 is a schematic diagram of the low-angle fracture logging response;

[0033] Figure 2 is a schematic diagram of the logging response of a medium-angle fracture;

[0034] Figure 3 is a schematic diagram of high-angle fracture logging response;

[0035] Figure 4 It is a flow chart of processing and interpretation of azimuthal conductivity simulation imaging method;

[0036] Figure 5 This is a top view of the six-arm formation dip logging instrument;

[0037] Figure 6 It is a diagram of the two-dimensional matrix method of the azimuthal conductivity around the well;

[0038] Figure 7 This is a comparison chart of the processing results of commonly used interpolation algorithms;

[0039] Figure 8 It is a diagram of the azimuthal conductivity two-dimensional matrix color-coded mapping method;

[0040] Fig. 9 This is a comparison of high-angle fracture response characteristics (well TZ1);

[0041] Fig.10 This is a comparison of the response characteristics of medium-angle fractures (well TZ1);

[0042] Fig.11 This is a comparison of low-angle fracture response characteristics (well TZ2);

[0043] Fig.12 This is a comparison diagram of induced fracture response characteristics (TZ3 well). DETAILED DESCRIPTION

[0044] The present invention discloses a fracture identification and evaluation method, such as Figure 4 As shown, the following steps are included:

[0045] Step 1: Set the window length and step length parameters, and read the azimuth angle of plate 1 and the conductivity data of plates 1 to 6 of the six-arm formation dip logging instrument; the window length parameter is set to 1.0-10.0 m, based on the fact that the fracture heights in the statistical work area are mostly below 1.0 m, and the maximum fracture height rarely exceeds 10.0 m. In this embodiment, it is set to 1.0 m; the step length parameter is set to 0.00254-0.001 m, based on the fact that the sampling interval of the dip logging instrument is 0.00254 m, and the vertical resolution of the dip logging conductivity is 0.001 m. In this embodiment, it is set to 0.00254 m.

[0046] Step 2: The azimuth of each adjacent plate of the six-arm formation dip logging instrument is 60° apart ( Figure 5), taking the azimuth angle of the No. 1 plate as the reference, the azimuth angles of the other plates are regularized within the range of 0° to 360°, that is, when the azimuth angle of any plate is greater than 360°, the azimuth angle of the plate is subtracted by 360°, so that the azimuth angles of each plate are between 0° and 360°;

[0047] Step 3: Construct a two-dimensional matrix M(x i ,y i ), first divide the 360° azimuth of a certain depth point within the window length into 6×m (requirement: m is a factor of 60 and m≥6) equal parts. In this embodiment, m=6, that is, it is divided into 36 equal parts. Use the azimuth of each plate to calculate the azimuth of each equally divided point as the matrix M(x i ,y i ) i Coordinates, and at the same time, the conductivity of each equally divided point is calculated by using the azimuth angle and conductivity data of each plate as interpolation sample points, as the matrix M(x i ,y i ) of y i coordinate( Figure 6 );

[0048] The interpolation method used in the interpolation calculation is selected from one of four interpolation methods: linear interpolation, cubic spline interpolation, cubic B-spline interpolation and Akima interpolation. In this embodiment, Akima interpolation is preferred.

[0049] The azimuth angle and conductivity data of plates 1 to 6 are used as sample points M(x i ,y i ), and compared the interpolation results of the conductivity data of the equally divided points around the well using four commonly used interpolation algorithms, including linear interpolation, cubic spline interpolation, cubic B-spline interpolation, and Akima interpolation ( Figure 7 ), the four interpolation methods can all meet the requirements of logging data interpolation and resampling, but the effects are not the same. The linear interpolation method has a simple calculation principle and a small amount of calculation, but the calculation result is only determined by the values ​​of adjacent interpolation sample points, and there are defects such as poor smoothness and unnatural curve transition; the cubic spline interpolation method has smooth calculation results, but it is easy to produce a large offset to the local extreme points, and the calculation result has poor stability; the cubic B-spline interpolation method is a global smooth fitting interpolation method, which avoids the problem of curve oscillation at local extreme points and is mainly suitable for steady-state data interpolation operations. Any data error in the interpolation sample will affect all calculation results, and the error transmission is strong; the Akima interpolation method is a local interpolation method with rigorous mathematical principles, simple calculation process, strong continuity of the calculation curve, high fitting accuracy and not prone to oscillation. Through comparative analysis, the Akima interpolation method is selected to interpolate the conductivity data of equally divided points around the well to obtain the conductivity data around the well, and construct a two-dimensional azimuthal conductivity matrix M(x i,y i ).

[0050] Determine whether the data within the window length depth range has been processed. If not, increase a step size, jump to the next sampling depth, and execute steps 2 to 3 until the azimuthal conductivity matrix within the window length depth range is fully constructed;

[0051] Step 4: Perform two-dimensional matrix color mapping to generate a simulated image. First, set any 2 to 5 color spectra with different RGB color modes, and map the two-dimensional matrix M(x i ,y i ) According to the conductivity value y i Arrange in order from small to large, when the conductivity values ​​are equal, according to the orientation value x i Arrange in order from small to large. When the conductivity value and the azimuth value are equal, they are sorted in the same order (sorting in the same order means that these equal matrix elements are treated as the same partition position, and correspond to the same color when performing chromatographic mapping later); then perform chromatographic partitioning to generate a simulated image of the azimuth conductivity around the well; the image coverage rate can reach 100%, which is significantly higher than the coverage rate of traditional electrical imaging logging images.

[0052] The color spectrum of this embodiment is set to four colors: white, yellow, red, and black, and is divided into three partitions: white-yellow, yellow-red, and red-black. The color spectrum partition process is: the matrix element at the first position in the ranking corresponds to white RGB (255:255:255), the matrix element at the third position in the ranking (if the position is not divisible, the position is determined by rounding) corresponds to yellow RGB (255:255:0), the matrix element at the second third position in the ranking (if the position is not divisible, the position is determined by rounding) corresponds to red RGB (255:0:0), and the matrix at the last position in the ranking corresponds to black RGB (0:0:0), thereby dividing the two-dimensional matrix into three color spectrum partitions;

[0053] The RGB color scale values ​​corresponding to the matrix elements of the remaining sorting positions are obtained by the mixed operation mapping method to realize the well circumference azimuthal conductivity two-dimensional matrix M(x i ,y i ) and the one-to-one mapping of all elements in the RGB mode color scale ( Figure 8 ), as follows:

[0054] ① Determine the multiplication factor K and addition factor A of the mixed operation mapping method: Assume that the minimum conductivity value within each chromatographic partition is E min , the maximum conductivity is E max , the left color scale value of each chromatographic partition is C lvalue , the right color value is C rvalue , when E min=E max When the multiplication factor K = 0, the addition factor A = C lvalue ; when E min ≠E max When , the multiplication factor K and the addition factor A are calculated using formula (1) and formula (2) respectively:

[0055]

[0056] A=C lvalue -K×E min (2)

[0057] ②Use the mixed operation mapping method to achieve a one-to-one correspondence between the two-dimensional matrix and the color label: Let the two-dimensional matrix M(x i ,y i ) color scale corresponding value is M RGB (x i ,y i ), then we have formula (3):

[0058] M RGB (x i ,y i )=M(x i ,y i )×K+A (3)

[0059] ③ After completing the two-dimensional matrix color code mapping of the first chromatographic partition, go to the second chromatographic partition and execute steps ① and ② until the two-dimensional matrix color code mapping within the range of all chromatographic partitions is completed.

[0060] Step 5: Image inspection and optimization, determine whether the image effect meets expectations. If not, check whether the parameter settings are reasonable, reset the parameters, and execute steps 1 to 5 to optimize the image until the effect meets expectations. The parameters that need to be checked include window length, step length, number of azimuths, and color spectrum. After that, determine whether the data of the entire well section has been processed. If not, increase the step length at a deeper level, jump to the next window length, and execute steps 1 to 5 until the data of the entire well section has been processed.

[0061] Step 6: Identify the fault type through image visualization and quantitatively calculate six parameters, namely, fault density, length, height, dip, strike, and inclination. Save the data and image and end the program.

[0062] According to the occurrence and genesis of the fractures, fractures are divided into four types: high-angle fractures, medium-angle fractures, low-angle fractures and induced fractures. In this embodiment, the response characteristics of the azimuthal conductivity simulation images of different types of fractures are summarized by comparing and analyzing the azimuthal conductivity simulation images of the present invention with the electrical imaging logging images. Based on the comparative analysis of the processing results of the well data in the work area, it is further confirmed that the identification and evaluation results of this method are reliable, and the accuracy of conventional logging data in identifying and evaluating fractures is effectively improved.

[0063] 1. Application example comparison

[0064] (1) Identification and evaluation of high-angle fractures

[0065] High-angle faults refer to faults with an inclination angle greater than 70°. This type of fault is mostly caused by tectonic action, with a wide opening and a large fracture height. It is easy to connect medium-angle faults and low-angle faults, thereby forming a fault network, which can greatly improve the permeability of the reservoir and has a significant destructive effect on the sealing of the dense barrier layer and cap rock. It is an important channel for long-distance migration of oil and gas. High-angle faults are often displayed as black "double tracks" with an interval of 180° on the electrical imaging logging map, which is approximately parallel to the well axis. Compared with the 192 sets of button measurement electrodes in imaging logging, the inclination logging has only six arm measurement electrodes, and the "blind area" between the plates is large. High-angle faults are mostly displayed as black "double tracks" with an interval of 180° and approximately parallel to the well axis on the azimuthal conductivity simulation imaging map. However, the "continuity" of the double track is weaker than that of the electrical imaging map. At the same time, since the data between the plates of the azimuthal conductivity simulation imaging method comes from interpolation operations, the width of the "double track" is greater than that of the electrical imaging map.

[0066] Comparison of the interpretation results of three methods for high-angle faults in the 3949-3953m section of Well TZ1 ( Fig. 9 ). The high-angle fractures in the electrical imaging map are displayed as black "double tracks" with an interval of 180°. Conventional logging methods such as conductivity anomaly detection, wellbore comparison, and conductivity comparison have no obvious response characteristics. The high-angle fractures in the azimuthal conductivity simulation image are displayed as black "double tracks" with an interval of 180°. The continuity of the "double tracks" is weaker than that of the electrical imaging map, and the width of the "double tracks" is greater than that of the electrical imaging map. The fracture density of the simulation image is the same as that of the electrical imaging map. The fracture length, height, and dip angle of the simulation image are relatively close to those of the electrical imaging map, and the relative error does not exceed 20%. The fracture strike and dip of the simulation image are basically the same as those of the electrical imaging map (Table 1).

[0067] Table 1 Comparison of high-angle fracture interpretation results

[0068]

[0069] (2) Identification and evaluation of medium-angle fractures

[0070] Medium-angle faults refer to faults with an inclination angle between 30° and 70°. This type of fault is mostly caused by tectonic action and often appears in groups. Affected by the pressure of the overlying formation, the fracture opening increases with the increase of the inclination angle, which can greatly improve the permeability of the reservoir and has a certain destructive effect on the dense shielding conditions. It is an important space for oil and gas migration and storage. Medium-angle faults are often displayed as black trajectory lines in the shape of sine or cosine on the electrical imaging logging map, and also as black trajectory lines in the shape of sine or cosine on the azimuthal conductivity simulation imaging map. Due to the limitation of the "blind area" measurement between the plates, the continuity of the black trajectory line on the simulation imaging map is weaker than that on the electrical imaging map, and the width of the black trajectory line is greater than that on the electrical imaging map.

[0071] Comparison of the interpretation results of three methods for medium-angle fractures in the 3880-3883 m section of Well TZ1 ( Fig.10 ). In the electrical imaging map, the equal-angle fractures are displayed as black cosine-shaped trajectory lines. Conventional methods such as conductivity anomaly detection, wellbore comparison, and conductivity comparison have no obvious response characteristics. In the azimuthal conductivity simulation imaging map, the equal-angle fractures are displayed as black cosine-shaped trajectory lines. The continuity of the trajectory line is weaker than that of the electrical imaging map, and the width of the trajectory line is larger than that of the electrical imaging map. The fracture density of the simulation imaging map is basically the same as that of the electrical imaging map. The fracture length, height, and dip angle of the simulation imaging map are basically the same as those of the electrical imaging map, and the relative error does not exceed 5.0%. The fracture strike and dip of the simulation imaging map are basically the same as those of the electrical imaging map (Table 2).

[0072] Table 2 Comparison of medium angle fracture interpretation results

[0073]

[0074]

[0075] (3) Identification and evaluation of low-angle fractures

[0076] Low-angle faults refer to faults with an inclination angle of less than 30°. This type of fault is mostly caused by tectonic action or differential diagenesis. The fault height is usually low and the opening is usually small, but the lateral continuity of the fault is strong. It is an important channel for controlling the lateral migration of oil and gas. Low-angle faults are often displayed as black trajectory lines with low amplitude sine, cosine or horizontal shapes on electrical imaging logging diagrams, and also as black trajectory lines with sine, cosine or horizontal shapes on azimuthal conductivity simulation imaging diagrams. Due to the limitation of the "blind area" measurement between the plates, the continuity of the black trajectory lines on the simulation imaging diagram is weaker than that on the electrical imaging diagram, and the width of the black trajectory lines is greater than that on the electrical imaging diagram.

[0077] Comparison of the interpretation results of three methods for two bottom angle faults in the 3894-3896m section of Well TZ2 ( Fig.11). Both low-angle fractures in the electrical imaging map are displayed as sinusoidal black trajectory lines. The conventional logging conductivity anomaly detection method has a certain conductivity anomaly display for the first low-angle fracture with a relatively large opening (the opening is 0.0354mm), but the intuitiveness is poor, and there is no obvious response to the second low-angle fracture with a relatively small opening (the opening is 0.0211mm). The two low-angle fractures on the azimuthal conductivity simulation image are displayed as sinusoidal black trajectory lines. The continuity of the trajectory line is weaker than that of the electrical imaging map, and the width of the trajectory line is larger than that of the electrical imaging map. The fracture density of the simulation image is the same as that of the electrical imaging map. The fracture length, height, and dip angle of the simulation image are relatively similar to those of the electrical imaging map, and the relative error does not exceed 20.0%. The fracture strike and dip of the simulation image are similar to those of the electrical imaging map (Table 3).

[0078] Table 3 Comparison of low-angle fracture interpretation results

[0079]

[0080] (4) Identification and evaluation of induced crack simulation images

[0081] Induced fractures refer to micro fractures formed during the drilling process due to changes in the original formation pressure or ground stress balance, resulting in pressure or stress difference between the wellbore and the formation. This type of fracture is usually low in height, short in length, and narrow in opening. Although the fracture density is large, it is mostly formed during the drilling process after oil and gas accumulation, and has little effect on oil and gas migration and accumulation. However, the direction of the induced fracture is parallel to the direction of the maximum principal stress of the formation. The induced fracture parameter information can play a certain guiding role in the optimization of reservoir fracturing operation parameters. Due to the small scale of the induced fracture, it is displayed in a non-standard sine and cosine style on the electrical imaging map, mostly as a series of black track lines with a nearly "goose-shaped" distribution. Due to the existence of the "blind area" measurement between the plates, the "goose-shaped" characteristics of the induced fractures on the azimuthal conductivity simulation image are weaker than those on the electrical imaging map. At the same time, for some induced fractures with too narrow openings (statistical fracture opening ≤ 0.0160 mm), there is no obvious response feature on the azimuthal conductivity simulation image.

[0082] Comparison of the interpretation results of three methods for induced fractures in the 3945-3949 m section of Well TZ3 ( Fig.12 ). The induced fractures in the electrical imaging are mostly displayed as black trajectory lines arranged in a "goose-like" pattern. The conductivity anomaly detection, wellbore comparison, conductivity comparison, etc. are mostly displayed in a messy manner, with strong multi-solution. The induced fractures in the azimuthal conductivity simulation imaging are displayed as black trajectory lines arranged in a nearly "goose-like" pattern. Due to the existence of the "blind area" between the conductivity logging plates, the regularity of the image is weaker than that of the electrical imaging. The parameters such as the density, length, height, density and inclination of the induced fractures calculated by the simulation imaging are all smaller than those of the electrical imaging, with a relative error of 20% to 50.0%. The direction and inclination of the induced fractures in the simulation imaging are relatively close to those in the electrical imaging (Table 4).

[0083] Table 4 Comparison of induced fracture interpretation results

[0084]

[0085] 2. Application effect analysis

[0086] The practice of processing and interpreting a large number of well logging data in the Ordos Basin shows that the azimuthal conductivity imaging recognition and evaluation method can intuitively identify four types of faults, namely high-angle faults, medium-angle faults, low-angle faults and induced fractures, by digital image processing of conventional well logging data, and can quantitatively evaluate six parameters, namely fracture density, length, height, strike, dip and inclination, especially the judgment of the strike and dip of various types of fractures is relatively accurate. Due to the limitation of the "blind area" of the plate measurement, this method has the highest accuracy in the recognition and evaluation of medium-angle fractures, and the average relative error of fracture length, height, density and dip is no more than 5.0%; the recognition and evaluation accuracy of high-angle fractures and low-angle fractures is second, and the average relative error of fracture length, height, density and dip is no more than 20.0%; the recognition and evaluation accuracy of induced fractures is relatively the lowest, and the average relative error of fracture length, height, density and dip is no more than 50.0% (Table 5). By increasing the number of acquisition plates and the measurement speed of the instrument, the "blind area" of the plate measurement can be reduced to a certain extent, which helps to improve the accuracy of the azimuthal conductivity imaging method. In general, the azimuthal conductivity imaging method combines the outstanding advantages of high acquisition rate of conventional logging data and strong visualization of imaging logging, which greatly improves the accuracy of visual identification of faults and quantitative evaluation of fault parameters using conventional logging data.

[0087] Table 5 Effect of azimuthal conductivity imaging method on identifying and evaluating fractures

[0088]

[0089] In summary, due to the complexity of fault genesis, the diversity of fault occurrence and the popularity of data, the existing fault identification and evaluation methods such as geological outcrop and core sample observation method, tracer dynamic test method and seismic data interpretation method have different degrees of application limitations. Scholars generally pay attention to conventional well logging machine learning algorithms and imaging logging visualization identification and evaluation of faults, but often ignore the research on methods for visual fault identification and evaluation using conventional well logging data. In view of the limitations of conventional well logging fault interpretation method such as poor visualization effect, difficulty in quantification, low imaging logging acquisition rate and white strip blind area in imaging map, the outstanding advantages of conventional well logging high acquisition rate and imaging logging strong visualization are integrated, and the conventional formation dip logging conductivity (or resistivity) and plate azimuth data are used to establish an azimuth conductivity simulation imaging fault identification and evaluation method. Compared with conventional methods, it has the advantages of visualization, quantification and high reliability; compared with imaging logging interpretation method, it has the advantages of universal survey of oil and gas field data, high imaging map coverage and low cost (Table 6), which has important promotion and application value.

[0090] Table 6 Comparison of application effects of fracture identification and evaluation methods

[0091]

[0092] The principle of the present invention is that the six-arm formation dip logging instrument is lifted from the bottom of the well in a rotary manner to measure the azimuth of the No. 1 plate (the angle between the No. 1 plate and the due north direction) and the conductivity (or resistivity) data of the No. 1 to No. 6 plates at different depths. When the formation develops a fracture, the drilling mud invades the fracture, which will cause the conductivity at the fracture to show abnormal changes, which is significantly different from the non-fracture development layer. When a low-angle fracture develops, the abnormal response connection line of the conductivity (or resistivity) curve of the No. 1 to No. 6 plates presents a nearly horizontal curve feature; when a medium-angle fracture develops, the abnormal response connection line of the conductivity (or resistivity) curve of the No. 1 to No. 6 plates presents a nearly sine or cosine curve feature; when a high-angle fracture or induced fracture develops, due to the diversity of the fracture direction and occurrence, there will be a variety of response characteristics such as non-polar plate detection, monopolar plate detection, adjacent bipolar plate detection, and spaced bipolar plate detection. Therefore, according to the response characteristics of the conductivity (or resistivity) curve, the fracture can be visually identified and the fracture parameters can be quantitatively evaluated.

Claims

1. A fracture identification and evaluation method, characterized in that: The following steps are involved: Step 1: Set the window length parameters and step length parameters, and read the azimuth of the No. 1 plate and the conductivity data of all plates of the multi-arm formation dip logging tool; Step 2: The azimuths of each adjacent plate of the multi-arm formation dip logging instrument are separated by 360° / n, where n represents the number of plates; taking the azimuth of plate No. 1 as a reference, the azimuths of the other plates are regularized within the range of 0° to 360°, that is, when the azimuth of any plate is greater than 360°, the azimuth of the plate is subtracted by 360°, so that the azimuths of each plate are between 0° and 360°; Step 3: Construct a two-dimensional matrix M(x i ,y i ), first divide the 360° azimuth of a certain depth point within the window length into m×n equal parts, where m represents the number of equal parts between two adjacent plates, and m≥n; use the azimuth of each plate to calculate the azimuth of each equally divided point as the matrix M(x i ,y i ) i Coordinates, and at the same time, the conductivity of each equally divided point is calculated by using the azimuth angle and conductivity data of each plate as interpolation sample points, as the matrix M(x i ,y i ) of y i coordinate; Determine whether the data within the window length depth range has been processed. If not, increase the depth by one step, jump to the next sampling depth, and execute steps 2 to 3 until the azimuthal conductivity matrix within the window length depth range is fully constructed; Step 4: Perform two-dimensional matrix color mapping to generate a simulated image. First, set any 2 to 5 color spectra with different RGB color modes, and map the two-dimensional matrix M(x i ,y i ) According to the conductivity value y i Arrange in order from small to large, when the conductivity values ​​are equal, according to the orientation value x i Arrange in order from small to large. When the conductivity value and the azimuth value are equal, they are sorted in the same order. Then, perform chromatographic partitioning to generate a simulated image of the azimuth conductivity around the well. Step 5: Image inspection and optimization, determine whether the image effect meets expectations. If not, check whether the parameter settings are reasonable, reset the parameters, and execute steps 1 to 5 to optimize the image until the effect meets expectations. The parameters that need to be checked include window length, step length, number of azimuths, and color spectrum. After that, determine whether the data of the entire well section has been processed. If not, increase the step length at a deeper level, jump to the next window length, and execute steps 1 to 5 until the data of the entire well section has been processed. Step 6: Identify the fault type through image visualization and quantitatively calculate six parameters, namely fault density, length, height, dip, strike, and inclination, and then save the data and images.

2. A fracture identification and evaluation method according to claim 1, characterized in that: In step 1, the multi-arm formation dip logging tool is selected from one of four-arm, six-arm, eight-arm, sixteen-arm and twenty-four-arm formation dip logging tools.

3. A fracture identification and evaluation method according to claim 1, characterized in that: In step 1, the window length parameter is 1.0 to 10.0 m; the step length parameter is 0.00254 to 0.001 m.

4. A fracture identification and evaluation method according to claim 1, characterized in that: In step three, the interpolation method used in the interpolation calculation is selected from one of four interpolation methods: linear interpolation, cubic spline interpolation, cubic B-spline interpolation and Akima interpolation.

5. A fracture identification and evaluation method according to claim 1, characterized in that: In step 4, the color spectrum is set to four colors: white, yellow, red, and black, and is divided into three partitions: white-yellow, yellow-red, and red-black.

6. A fracture identification and evaluation method according to claim 5, characterized in that: In step 4, the chromatogram partitioning process is as follows: the first-ranked matrix element corresponds to white RGB (255:255:255), the third-ranked matrix element corresponds to yellow RGB (255:255:0), the second-third-ranked matrix element corresponds to red RGB (255:0:0), and the last-ranked matrix element corresponds to black RGB (0:0:0), thereby dividing the two-dimensional matrix into three chromatogram partitions; The RGB color scale values ​​corresponding to the matrix elements of the remaining sorting positions are obtained by the mixed operation mapping method to realize the well circumference azimuthal conductivity two-dimensional matrix M(x i ,y i ) is mapped one by one to the RGB mode color scale, as follows: ① Determine the multiplication factor K and addition factor A of the mixed operation mapping method: Assume that the minimum conductivity value within each chromatographic partition is E min , the maximum conductivity is E max , the left color scale value of each chromatographic partition is C lvalue , the right color value is C rvalue , when E min =E max When the multiplication factor K = 0, the addition factor A = C lvalue ; when E min ≠E max When , the multiplication factor K and the addition factor A are calculated using formula (1) and formula (2) respectively: A=C lvalue -K×E min (2) ②Use the mixed operation mapping method to achieve a one-to-one correspondence between the two-dimensional matrix and the color label: Let the two-dimensional matrix M(x i ,y i ) color scale corresponding value is M RGB (x i ,y i ), then we have formula (3): M RGB (x i ,y i )=M(x i ,y i )×K+A (3) ③ After completing the two-dimensional matrix color code mapping of the first chromatographic partition, go to the second chromatographic partition and execute steps ① and ② until the two-dimensional matrix color code mapping within the range of all chromatographic partitions is completed.

Images